This research examines how different categorizations of immigrant generations under- or over-estimate their socioeconomic assimilation outcomes such as educational attainment, earnings, and health-related outcomes.  Ethnic stratification researchers, for example, seem to have lumped the 1.5 and the first immigrant generations together mainly because of methodological convenience rather than for reasons based on theoretical or empirical considerations and have significantly underestimated the socioeconomic disadvantage of first-generation immigrants.  This paper pays a careful attention to categorization of immigrant generations based on comparative analyses of data from both American Community Surveys (ACS) and Current Population Surveys (CPS) by focusing on Hispanic and Asian immigrant subgroups’ assimilation outcomes. CPS is the best nationally-representative data set for investigating how immigrant assimilation outcomes vary by immigrant generation thanks to the inclusion of questions on parental birthplace. However, it suffers from a small sample size, particularly for small ethnic groups, compared with ACS.  While ACS currently does not allow researchers to precisely measure immigrant generation, it has been widely used to measure immigrant assimilation outcomes because of its large sample size and detailed ethnic group categories. Findings on estimation issues between the two data sets will yield significant implications in evaluating the long-term incorporation of immigrants.